For medical illustrators and animators, the 3D reconstruction function of OsiriX is a very useful tool. Using OsiriX, 3D renderings from DICOM datasets can be used to view anatomical structures as a reference for 2D illustrations and 3D animations. A 3D volume rendering in OsiriX can be used to dynamically visualize anatomical structures in any view that a medical illustrator may need for an illustration. A 3D surface rendering can be exported from OsiriX as an .obj file and can be imported into a 3D modeling and animation program like 3DS Max to be used as a guide to create a new 3D model (see tutorial) .
Looking at the DICOM sample image sets from the OsiriX website, it was difficult for me to choose which type of scans to use for reference, and once chosen, to decide which series in the study to use. In addition, I was not familiar with the radiology terms used to describe the datasets.
Last week, at the Frank Armitage lecture at the University of Illinois at Chicago , I met Dr. Jeremy Durack, an Assistant Professor of Clinical Radiology at the University of California , San Francisco . He is also an interventional radiologist and researcher and works with Dr. Paul Brown, President of eHuman . Dr. Durack answered some of my questions regarding the types of scans to use to create 3D renders in OsiriX. He also cleared up some radiology terms for me.
osiriximaging : In OsiriX, what is the best type of scan to use for a 3D rendering such as a 3D surface rendering, 3D volume rendering, 3D MIP? What is the worst type of scan to use?
Dr. Durack : High resolution volumetric imaging studies are ideal for 3D rendering using software such as Osirix. Computed tomography (CT) scan datasets produce tremendous 3D renderings. While magnetic resonance imaging (MRI) is a cross-sectioning imaging technique like CT, gaps between images often degrade the quality of 3D renderings. Of note, 3D MR sequences can be performed, though these sequences require longer scan times. MR imaging is less well suited to current software used for rendering as the software is usually developed for CT Hounsfield unit transformation (density histograms), whereas MR imaging is based on inherent tissue properties other than tissue density. Other modalities, such as PET scanning or ultrasonography don’t provide the resolution necessary to accurately render in 3D. Conventional x-ray or fluoroscopic studies (including angiography) are 2D projections, therefore are unable to be used for volumetric rendering.
osiriximaging : Why would a radiologist use a 3D MIP over a 3D volume rendering? What are the advantages/ disadvantages of both?
Dr. Durack : MIP (maximum intensity projection) are used primarily to better delineate vascular structures during angiographic studies. MIPs show all of the vessels along a particular projection (usually stand out due to intravascular contrast increasing their density), but vessel outlines merge making it very difficult to determine the spatial relationships between vessels. 3D volume rendered images allow one to better visualize spatial relationships between structures, particularly when the perspective can be changed enabling rotation of the line of sight around them.
osiriximaging : In OsiriX, what type of scan is best to create a 3D render of muscles, arteries, nerves, organs, tumors and bones?
Dr. Durack :
muscles: CT
arteries: Contrast enhanced CT obtained in the arterial phase of contrast injection
nerves: CT
organs: Multi-phase contrast enhanced CT
tumors: Contrast enhanced CT, obtained in multiple contrast phases (arterial, portal-venous, venous phases - depending upon tumor type and location).
bones: CT
Of note, muscles and nerves can be difficult to render from any imaging study as they do not enhance with contrast the way many other vascular organs do. I selected CT for those structures simply because it is the highest resolution volumetric study and with proper manipulation of the rendering transfer function, they can be teased out from surrounding tissues despite minimal tissue density differences.
osiriximaging : What is the difference between a CT, CT 16, CT 16/64, and CT 64?
Dr. Durack : Each number refers to the number of slices obtained in each rotation of the CT scanner. Early CT scanners were single-slice, thus each rotation resulted in single imaged slice. As CT technology has developed, newer generations of CT scanners can obtain multiple slices in a single rotation (4, 8, 16, 32, 64, and so on). The more slices obtained per rotation, the faster the scan acquisition time.
osiriximaging : What is the difference between a PET/CT and PET/CT 64?
Dr. Durack : Again, the number simply refers to the number of slices per rotation. A PET/CT scanner combines the ability to acquire positron emission tomographic images and computed tomographic images from the same machine. Images from each modality are registered to allow the radiologist to evaluate the relationship between physiologic processes (such as tumor metabolism) and anatomic structures.
osiriximaging : What is a cardiac gated PET and CT?
Dr. Durack : Electrocardiographic data is obtained during a scan in order to relate the PET or CT data to the cardiac cycle. Images can be acquired throughout the heart cycle and retrospectively registered according to specific points in the cycle, or imaging can be obtained only during a specific phase of the cardiac cycle.
osiriximaging : What is an extremities runoff?
Dr. Durack : “Runoff” studies are studies obtained for evaluation of the arteries in the legs. Contrast is injected into a peripheral vein (most commonly) in the case of CT runoff studies or directly into the aorta for conventional angiography. When the contrast is injected into a peripheral vein, the CT imaging is held up until the contrast reaches the arteries supplying the legs, then images are quickly obtained as the contrast flows through the arteries. The images are acquired as the contrast ‘runs off’ into the legs so that the arteries (or veins) of the extremities can be visualized. This technique is valuable for assessing blockages to arterial blood flow or narrowing in arteries, even in small blood vessels. Without a dense contrast agent to highlight the blood vessels, they would be very difficult to distinguish from surrounding tissues of similar density.
osiriximaging : What is a 3T scanner?
Dr. Durack : T stands for Tesla. The number 3 indicates how many Tesla the magnet in the MRI machine is capable of generating. In general, higher Tesla magnets (the stronger the magnet), produce higher resolution imaging with better tissue differentiation.
Esther Pulley is a recent graduate of the Biomedical Visualization program at the University of Illinois at Chicago.
How should I request a CT to the radiology dept. in order to have a proper CD for later Osirix working of angio-CT’s? (slice thickness, number of images, any other tech. details)